After someone posted a Steam Turbines based Salt Water Boiler, I wondered if there was a way to boil salt water and get water back without Steam Turbines, and hopefully at a better power efficiency. After running the numbers, it turns out there is! I present to you the most power-efficient way to turn salt water (yes, more efficient than the Desalinator too!) before Space materials, which doesn't require Dupe operation, and is one of the most compact save from Desalinator.
More maths at the end, and without further ado, the build:
Obviously, the top left part is where the Salt Water Geyser you are taming is supposed to be. This build works with 95°C Salt Water, and has not been tested for other temperatures. It will work with any Salt Water Geyser, regardless of its erupting output. I tested it with a constant 2kg/s, 5kg/s and 8kg/s Salt Water.
The Salt Water enters on the left of the 3 Mechanized Doors, until 20kg accumulates on the tile right of the door. At that point, the 3 doors close in sequence, to avoid mass deletion. Note that both tiles right of the rightmost door must be free. Early designs which had a tile on the top one resulted in a 10% mass deletion (thanks to Hotep for figuring that out).
The Salt Water sits above of the AT, which is submerged in 4 tiles of Crude Oil (or any other liquid that stays liquid within 95°C <=> 110°C and is heavier than Salt Water). This part is important to only heat Salt Water and not heat Steam, which would lower the efficiency.
Once the chamber is filled, the AT is turned on to boil the salt water into steam. Steam will naturally expand to the right and be condensed by the Radiant Liquid Pipes into 95°C Water, and be pumped to wherever you need water (to compare with ST-based builds, in which the STs act as a pump). That pump is controlled by an above 20kg Hydro Sensor to avoid pumping partial packets.
There is another Hydro Sensor at less than 200kg (or any other value of your choosing over about 100kg) to stop the boiler in case the water output is clogged. This will in turn stop the AT, which will stop the boiling, then the door will stay closed, and eventually your Salt Water Geyser will over-pressure and safely stop.
There is also a Salt Sweeper, linked to Pressure Plates to lower the power consumption. Timer/cycle sensor would work just as well if you want to save 2 tiles at the bottom.
A few important details: when flowing, the Salt Water is blocked by a tile on the right, to ensure it doesn't fall in the clean Water tank. With the Hydro Sensor controlling doors, that has never happened. Between the Aqautuner and clean Water chamber, there is a 2 tiles height gap: this is important, because otherwise, on occasion, Water can condense on the ledge: if there is only 1 tile height, this drop of water will act as a liquid airlock and Steam will be trapped above the Aquatuner. While it can be solved with a Hydro-Sensor, this means sacrificing the Sweeper. To limit that occurrence, Radiant Liquid Pipes don't run close to the ledge.
As for materials, everything insulated is made out of Ceramic (I didn't test with Igneous Rock, it will probably work as well). The Pump must be made out of Gold Amalgam or better to handle 95°C water. The Sweeper and Loader must be made out of Iron, Copper, Gold or better to handle the Steam temperature. The doors and Themo-Aquatuner can be made out of literally anything. The test build use Steel door (subpar because of high TC) and Gold Amalgam AT, but since the AT never goes over 125°C, even Copper Ore will work. The Radiant
The cooling loop use plain Water in the test build, but can use any Water type. If you are late in the game Super coolant will almost double the power efficiency, but that transforms an otherwise early-game build in a late-game build. I only performed tests with plain water, but it should work just as well with Super coolant. Note that at that point, you are better off using STs.
It turns out that, without any special mechanism, the system will balance itself: 95°C Salt Water comes in, 95°C Water goes out, and the cooling runs in a closed loop.
Note that since everything will generally be colder before starting the build, you will need to dump cold coolant and replace it with hot coolant until the system is up to operating temperature. Once it's at operating temperature (it shouldn't take more than a cycle), you can close the loop.
Let's delve into the maths. All numbers are for 5kg/s Salt Water at 95°C.
The desalinator uses 480W and is the reference.
The ST-less boiler (this build) need to heat 95°C Salt Water to 102°C to vaporize it, which translates to 143.5KDTU/s. Using Water, this means 24.5% Aquatuner uptime, or 294W. The pump will need to pump 4.65kg/s of water, resulting in 112W. The total is 406W, 15% better than the Desalinator in addition to not using Dupe labor.
With a ST-based boiler, without Split Turbine, you need to heat 95°C to 125°C Steam, which translates to 615KDTU/s. Using Water, this means 105% Aquatuner uptime, or 1260W. Using the 0.969W/KDTU/s figure, you get 596W back, for a net use of 664W, which is 38% worse than the Desalinator.
How about ST boiler using Split Turbine? It will be somewhere in the middle, but no matter what is going to be less efficient than a ST-less boiler, in addition to being the most complicated to build, for no practical advantage.
With Super coolant, a ST-less boiler will use only 146W of Aquatuner (12% uptime) for a total of 258W. A ST-based boiler will use 625W of Aquatuner (52% uptime), for a net use of 29W. Super coolant efficiency make it worth switching to STs because of the pump power draw in the ST-less boiler.
In practice, I tested my design for fixed inputs of 2kg/s, 5kg/s and 8kg/s. Each test was done for 5 cycles on the dot at 3x speed, controlled with a timer sensor. I got the results by multiplying buildings power draw with their "Last 5 cycles" uptime.
With all tests, the power draw is within 5% of the theory, and both Salt and Water outputs are within 1% what they should be, which means there is no mass deletion going on.
